dudaniavinash.github.io

Welcome to Avinash Dudani’s Portfolio !

About Avinash:!

  1. Avinash is an autonomous Industrial engineer with 3+ years of experience as a Data Analyst.
  2. He is innately curious and immensely passionate about Big Data, Cloud Computing, Business Intelligence, Process improvement, and Project Management.
  3. Avinash strives to create unqiue end-to-end durable, scalable and available solutions using products and services in the Microsoft Tech stack, providing speed to insights while promoting data integrity.
  4. He is proficient in utilizing tools such as Power BI, Azure, SQL Server Management Studio, Power Automate and Power Apps.
  5. Languages: Power Query M, DAX, Python, T-SQL.

Sample Projects

Project 1 : Adventure Works Sales Analysis

Problem Statement

  1. Analyze historical sales and profit and territory.
  2. Identify bestselling products and attributed customers and regions.
  3. Forecast revenue for the next 7 periods.
  4. Execute pricing scenario analysis to understand increase in product cost.

Tools : Power BI Skills: Data Analysis, Data Visualization and Storytelling.

Solution

Here is a video, demonstrating the functionalities below:

  1. Conditional drilldown using 2 columns: Customer ad Region to analyze Profit.
  2. Dynamic Product Pricing Scenario Analysis.
  3. Forecast Revenue for next 7 months+.
  4. Pareto Chart for Regions and Product Sub Categories. Using the 80-20 Principle to understand best sellers.
  5. Drill down Territories By Fiscal Year, Customer Details and (Sub)/Product Categories.
  6. Page Navigation and Custom Filter Panes.

A summary of this project is also available here.

Project 2 : Linear Regression

Problem Statement

Multiple linear regression to predict Profit based on Administration spend, Marketing spends, R&D spend and State

Tools: Python. Skills: Exploratory Data Analysis, Descriptive and Predictive Analysis.

Solution Steps:

  1. Importing Libraries such as sklearn,pandas, numpy and matplotlib. Reading csv files.
  2. Descriptive Analysis ( Exploring avaialble features,null counts and data types, descriptive statistics such as mean, min/max, IQR values.
  3. Exploratory Data Analysis
  4. Model Splitting, Training and Testing : Used a 80-20 Split
  5. Model Evaluation : On tuning the model, the R^2^ Score = 93.47%